The Rise of Computer Vision Technology

A lot of factors have contributed to the revolutionizing success of AI. Computer Vision is one of those driving elements. It is a sequential integration of three distinct processes, i.e. acquisition of images or visual stimuli from the real world in the form of binary data, image processing in form of edge detection, segmentation matching and lastly analysis and interpretation. From augmented reality games to self-driving cars to Apple's Facial Unlock feature, it has deeply impacted our life. And this influence is not free of consequences. However, on the flip side, it has been welcomed with generally encourage reviews.

The surveillance sector mainly benefits from implementing it. The most recent example is the detection of COVID-19 by studying body temperatures. CCTVs are also being used to monitor movements in public places, suspicious activities in areas like banks, laboratories, etc. or checking in if your pet is ruining your sofa while you are not at home or possible intruders. It will also make finding missing persons, scene reconstruction easier and enable faster security checkups at the airport.

Earlier if you wished to have a jacket of a particular style, you had to visit a number of shops looking for that perfect item, even if you were lucky, sometimes it turned out that they didn't have it available in your size. However, now, thanks to e-commerce sites, one can scan through an apparently infinite number of items of different designs, sizes, color, etc. with a click away. Sites like Amazon, Wish, manage shelves, personalize incentives, and calculate a shopper's bill.

Photo editing Apps like Abode Photoshop, Lightroom CC or Google Snapseed, etc., and Smartphones like Google Pixel, OnePlus, Realme, etc. are using computer vision for image recognition and restoration by de-noising and segment processing.  Phones now are giving DSLRs run for their money by being equipped with powerful processors that let them capture high-resolution photos in different manual settings. These photos can be zoomed to a greater extent without pixelating. Google Photos app lets you tag by location, categorize your photos on the basis of different tags like, animals, landscape and Scan an entity and provide you with details.  Apps like WeHeartIt and Pinterest show posts based on your searched items in your feed. Reverse imaging helps to look for photos' source and copyright infringement. Drones can carry out topographical modeling and agricultural mapping. It can also detect the presence of enemy or enemy's military deployments and missiles in the patrolled areas.

Have you been giving attendance proxies for your BFF at university? Futuristic Facial Biometrics can strip your friend of this luxury. Biometrics have been widely used in a variety of applications and purposes and will also be on the rise in the upcoming years. Fingerprint scanning for entry to Facial scanning before entrance exams, retina scanning for Aadhar card we have all been there. Even Facebook uses this image scanning to auto-detect you in family photos or Saturday Friends' Getaway. Although it may seem harmless, there has huge controversy regarding this. Either it is constant or occasional screening, it is argued that such technology invades personal space and makes them vulnerable. Instances, where AI is used for identification of gender or racial profiling etc., have been questioned and condemned. It is ruled out that due to the frequency of large-scale data harvesting there is a chance for an inaccurate report. There are evidences where governments around the world have been exploiting this to scrutinize people, especially during election campaigns, protests rallies and riots. In past, China used this to track people Islamic Minority people of Uyghur.

Face recognition software that is able to identify people even when they're wearing surgical masks, as the company Hanwang has developed, could also be used to identify people who obscure their faces at political protests out of fear of retribution from the government.

Surprisingly, there was a silver lining when recently Google removed the ability for users of its cloud services to label images with "man" or "woman" while training computer vision models.

In the retail sector, now buyers can opt for cashless payment by scanning QR codes and reduce the occurrence of shoplifting. Waiting in the queue to cast votes, withdraw money at ATM's or banks shall soon be ancient history. Computer Vision can foster consumer loyalty due to customizable personal experience. In Hospitals, medical personnel use computer vision to predict various types of cancer by examining x-rays and MRI scans. Google's computer vision, AI was able to detect false positives and negatives in diagnosing diabetic retinopathy 90 percent of the time, versus 80 percent with human doctors.

In manufacturing industries and overhauls, it facilitates auto – inspection and quality checks with human intervention. Autonomous cars use motion recognition and deep learning algorithms analyze video feeds from cameras and sensors installed on the vehicle to detect prevailing road conditions and obstacles like people, rocks, speed breakers, trees and navigate accordingly. Despite being endowed with these features, unlike humans, AI fails to practically use the information and knowledge to deal with situations that are not in their software codes. E.g., it cannot differentiate between a tasteful documentary about genocide and content or post that promotes violence. It cannot park vehicles in odd spaces etc.

Another major issue is processing data on a real-time basis and storage. Even in face-recognition technology, a minute error can lead to law enforcement against an innocent person.

Although the benefits outweigh the negative aspects of computer vision, one cannot choose to ignore them either. The future scope of this lies in working on them, eliminating error, and provide larger data while integrating the available resources. This is mandatory to augment pros and be in correspondence to human rights.